Specific language impairment (SLI) affects over 7% of US children and is associated with increased risk of academic and work-related problems. A key feature of SLI is difficulty learning new words; yet the nature of this difficulty is poorly understood. Both phonological memory and semantic processing deficits have been proposed to account for word learning problems in SLI. While there is some evidence for each proposal, previous studies have yielded mixed results. Given that the two proposals may suggest different approaches to assessment and intervention, there is a critical need to understand the relative contributions of phonological and semantic deficits to word learning in SLI. Thus, an overarching aim of our research program is to clarify word learning profiles in SLI, and to relate these profiles to other language and cognitive profiles, including phonological memory and semantic processing skills and deficits. One reason for variability in past findings may be the use of heterogeneous samples. Many - but not all - children with SLI also have dyslexia (SLI-DYS). These children have deficits in phonological memory, as measured by nonword repetition tasks, as well as semantic deficits. However, there is also evidence that children with SLI in the absence of dyslexia (SLI-only) have intact nonword repetition. Past studies have tended to conflate the two samples. To address this issue, we will study phonological and semantic contributions to word learning in four subgroups: SLI-only, DYS-only, SLI-DYS, and typically developing (TD) children (Specific Aim 1). We hypothesize the word-learning difficulties in SLI and dyslexia have different underlying causes. We will test this hypothesis through careful examination of word learning processes in the four groups using standardized assessment, cognitive-behavioral tasks, and event-related potentials (ERP). In addition, we will determine how well word learning performance in grade 2 predicts growth in vocabulary knowledge and comprehension skills one year later (Specific Aim 2). The transition from grades 2 to 3 marks an important turning point in vocabulary and comprehension development. We predict that word learning performance in grade 2 will account for unique variance in vocabulary growth and comprehension in grade 3. We further predict that semantic skills, in particular, will help predict outcomes among SLI-only children. This project is innovative three ways: 1) in the separation of phonological and semantic aspects of word learning in SLI; 2) in the separation of SLI-only and SLI-DYS samples; and (3) in the use of real-time ERP measures, which complement offline tasks. The long-term goal is to develop more effective assessments and better-targeted interventions for SLI, which would have a strong positive impact on public health outcomes.